Achievements and limitations of contemporary common bean breeding using conventional and molecular approaches

Common bean (Phaseolus vulgaris L.) improvement programs have been successful using conventional breeding methods to accomplish a wide array of important objectives. Specific achievements include the extension of range of adaptation of the crop, the development of cultivars with enhanced levels of disease and pest resistance and breeding lines that possess greater tolerance to drought. The most effective breeding method depends on the expression and inheritance of the trait to be selected and the target environment. Many bean improvement programs use molecular markers to facilitate cultivar development. In fact, several recent germplasm releases have used molecular markers to introgress and or pyramid major genes and QTL for disease resistance. Related species (P. coccineus and P. acultifolius) via interspecific hybridizations remain an important albeit long-term source for resistance to economically important diseases. Slow progress has been made in the improvement of traits such as adaptation to low soil fertility and tolerance to high levels of soluble Al in the soil using conventional breeding methods. The inability to directly measure root traits and the importance of genotype × environment interaction complicate the selection of these traits. In addition, symbiotic relationships with Rhizobium and mycorrhiza need to be taken into consideration when selecting for enhanced biological N fixation and greater or more efficient acquisition of soil P. Genomic examination of complex traits such as these should help bean breeders devise more effective selection strategies. As integration of genomics in plant breeding advances, the challenge will be to develop molecular tools that also benefit breeding programs in developing countries. Transgenic breeding methods for bean improvement are not well defined, nor efficient, as beans are recalcitrant to regeneration from cell cultures. Moreover, if issues related to consumer acceptance of GMOs cannot be resolved, traits such as herbicide tolerance in transgenic bean cultivars which would help farmers reduce production costs and decrease soil erosion will remain unrealized.     

Characterization of resistance to angular leaf spot of common bean (Phaseolus vulgaris L.) breeding line SPS50HB

The common bean (Phaseolus vulgaris L.) makes an important contribution to the human diet, particularly in Africa and Latin America. Because angular leaf spot (ALS), caused by the fungal pathogen Pseudocercospora griseola, is one of the most severe foliar diseases of the bean plant, an important priority is to identify genes encoding resistance. The present study focused on the resistance shown by the Mesoamerican common bean breeding line SPS50HB. From the pattern of segregation for resistance displayed in an F₂ population bred from a cross between SPS50HB and the ALS-susceptible Ethiopian variety Red Wolaita, it was concluded that the resistance of SPS50HB is controlled by two unlinked dominant genes. An allelism test indicated that one of these genes was either identical with, allelic to, or closely linked to the major gene Phg-2 carried by variety Mexico 54. The sequence-characterized amplified region assays OPEO4 and PF13, which are diagnostic for an ALS resistance gene carried by the germplasm accession G10909, both tracked a possible second gene present in SPS50HB.     

Pathogenic variation in, sources of, and breeding for resistance to Phaeoisariopsis griseola causing angular leaf spot in common bean

If we are to breed common bean (Phaseolus vulgaris L.) for durable resistance to diseases, we must understand pathogenic variation and find sources of resistance. Our first objective was to determine the patterns of pathogenic variation found among isolates of Phaeoisariopsis griseola (PG), the fungus that causes angular leaf spot (ALS) in common bean. We characterized 433 PG isolates from 11 Latin American and 10 African countries, using differential cultivars, isozymes, and/or random amplified polymorphic DNA (RAPD) markers. We also systematically screened, for ALS resistance, common bean accessions from the world collection held at CIAT, and assessed the progress so far made in breeding for resistance to ALS. Despite their great diversity within and between countries on both continents, the PG isolates were classified into two major groups: Andean, and Middle American. Although each group had internal differences for virulence, and biochemical and molecular characteristics, the ‘Andean’ PG isolates were more virulent on common beans of Andean origin, than on those of Middle American origin, thus, suggesting a host-pathogen co-evolution. The ‘Middle American’ PG isolates, although more virulent on common beans from Middle America, also attacked Andean beans, thus, exhibiting a much broader virulence spectrum. To find sources of resistance, we tested 22,832 common bean accessions against naturally occurring PG isolates in the field at CIAT's Experiment Station, Quilichao, Colombia, between 1985 and 1992. The resulting 123 intermediate (scores of 4 to 6) and resistant (scores of 1 to 3) accessions were then tested in the greenhouse against selected 14 PG isolates of diverse origins. Nineteen accessions were intermediate or resistant to at least 13 of 14 PG isolates. Similarly, of 13,219 bred lines tested in the field between 1978 and 1996, 89 were intermediate or resistant. Of these, 33 bred lines proved intermediate or resistant to at least eight of nine PG isolates to which they were challenged in the greenhouse. We suggest that, to breed for durable resistance to ALS, common bean populations should be developed from crosses between Andean and Middle American gene pools. The populations should then be systematically evaluated and selected against the broadest range of the most virulent PG isolates of diverse evolutionary origins. This revised version was published online in August 2006 with corrections to the Cover Date.     

Genomics,genetics and breeding of tropical legumes for better livelihoods of smallholder farmers

Legumes are important components of sustainable agricultural production, food, nutrition and income systems of developing countries. In spite of their importance, legume crop production is challenged by a number of biotic (diseases and pests) and abiotic stresses (heat, frost, drought and salinity), edaphic factors (associated with soil nutrient deficits) and policy issues (where less emphasis is put on legumes compared to priority starchy staples). Significant research and development work have been done in the past decade on important grain legumes through collaborative bilateral and multilateral projects as well as the CGIAR Research Program on Grain Legumes (CRP‐GL). Through these initiatives, genomic resources and genomic tools such as draft genome sequence, resequencing data, large‐scale genomewide markers, dense genetic maps, quantitative trait loci (QTLs) and diagnostic markers have been developed for further use in multiple genetic and breeding applications. Also, these mega‐initiatives facilitated release of a number of new varieties and also dissemination of on‐the‐shelf varieties to the farmers. More efforts are needed to enhance genetic gains by reducing the time required in cultivar development through integration of genomics‐assisted breeding approaches and rapid generation advancement.     

Development of common bean (Phaseolus vulgaris L.) lines resistant to the bean pod weevil,Apion godmani Wagner,in Central America

Summary The larva of the bean pod weevil (BPW), Apion godmani Wagner (Coleoptera: Curculionidae), causes serious yield losses in common bean (Phaseolus vulgaris L.) in Mexico and Central America, by consuming the seed as it develops in the immature pod. Resistance to the BPW was identified in bean germplasm of highland Mexican origin, and these sources of resistance were incorporated into a pedigree breeding program to recover locally adapted lines resistant to Bean Common Mosaic Virus and BPW, with commercial grain for Guatemala, Honduras and El Salvador. These lines yielded as well as or better than local cultivars in the absence of the insect, and better than local cultivars when the BPW was present. Resistance appeared to be governed by several genes, and was stable across geographic areas, seasons and planting systems.     

Evaluation of a DNA probe for the quantitative detection of common bacterial blight in common bean and its application in a breeding program

Summary Breeding of Phaseolus vulgaris L. for resistance to common bacterial blight (CBB) can be done with visual evaluations of symptoms to distinguish broad resistance classes, but a more quantitative measure was needed for genetic studies of resistance. A novel method of evaluation was developed by quantifying Xanthomonas campestris pv. phaseoli (XCP) in bean leaf tissue infected with CBB using a ³²P-labeled probe and densitometric analysis of hybridization signals. Quantification of bacterial populations using the probe was highly correlated (r=0.98) with the number of colony forming units (CFU) from plate counts of the same leaf samples. The probe was used to follow XCP population dynamics on susceptible (BAT 41) and resistant (OAC 88-1) bean genotypes. OAC 88-1 supported a maximum XCP population which was approximately tenfold less than BAT 41. The probe was also used to study an F₂/F₃ population segregating for resistance. Narrow sense heritability estimates were less for resistance measured on the basis of bacterial populations (0.18–0.26) than on visual scores of symptoms (0.29–0.38). The anticipated response to selection for CBB resistance would be less based on bacterial numbers than based on symptom expression in this population. In breeding for resistance to CBB, selection based on visual symptoms combined with measurements of XCP populations using a DNA probe can be used to develop bean genotypes that are both resistant to symptom development and bacterial multiplication.Abbreviations CBB common bacterial blight - CFU colony forming units - XCP Xanthomonas campestris pv. phaseoli     

Growth analysis in a reduced leaf mutant of common bean (Phaseolus vulgaris L.)

Summary A comparison between the Royal Red bean variety and a reduced leaf mutant on the basis of growth indices is reported. The mutants has lower leaf area index (LAI) and leaf area duration (LAD) (–43.7%), higher specific leaf weight (SLW) (on the mean +33.3%) and accumulates less dry matter (–7.7%) than the normal genotype. When the net assimilation rate (NAR) is considered, the mutant is superior to the normal (+62.3%). It is concluded that the unit area of the mutant leaves accumulates dry matter more efficiently than that of the corresponding normal variety.     

Development of an appropriate breeding scheme for tolerance to Empoasca kraemeri in common bean

Summary Significant advances in increasing tolerance to the leafhopper Empoasca kraemeri Ross & Moore in common bean have been obtained using a new breeding scheme where yield under leafhopper attack is the principal selection criterion in the evaluation of progenies. However, to further refine this breeding scheme, a study was conducted to determine whether selection for nonprotected yield would be more effective in early versus late generations. Two selection strategies were compared. In Strategy I, early generation selections in the F₂ and F₃ were compared to Strategy II where late generation selections were made in the F₄ and F₅, with the F₂ and F₃ generations advanced using single pod descent and bulk practices, respectively. Yield trials of the F₆ lines from both selection strategies were conducted under nonprotected and insecticide protected treatments. No significant differences were detected between the two selection strategies. However, Strategy II did produce advanced lines with greater nonprotected yields than did Strategy I, with the best F₆ line, in three out of four crosses, coming out of Strategy II. Late generation selection is recommended over early generation selection. Results of the new leafhopper breeding scheme, based on yield, are compared to the old breeding scheme where selections were made using visual selection practices in early generations.     

Sources and inheritance of earliness in tropically adapted indeterminate common bean

Summary Nineteen early maturing, indeterminate common bean parents were crossed with a tester line of normal maturity. All parents, F₁ hybrids, F₂'s, and backcross generations were evaluated at CIAT-Palmira, Colombia. All 19F₁ populations flowered and matured almost as early as the respective early parent (mean of 61 vs 59 days). The F₂ populations were slightly less early (63 days), but were still earlier than the mean of the two parents (65 days). Additive effects for days to maturity were usually larger than dominance or interaction effects. Narrow sense heritability for days to flower and maturity were high (greater than 0.67), while heritabilities of absolute and relative duration of pod filling were low (0.17 to 0.50). Also, our data indicated that for each day of increase in earliness, yield potential will decrease 74 kg/ha.     

Genetic diversity in pod and seed quality traits of common bean populations from northwestern Spain

Common bean (Phaseolus vulgaris L.) is an important food in the northwest of Spain where it is consumed as either green pods or dry seeds after cooking. Quality properties play a major role in cultivar acceptance by processors and consumers. To determine if sufficient variation in important culinary and nutritional traits exists within common bean, fifty-nine populations (plus five commercial cultivars) were grown in the northwest of the Iberian Peninsula in three different environments and evaluated for sixteen fresh pod and dry seed quality traits. Populations showed significant differences for pod curvature, length/width and width/thickness ratios of pod and seed, pod and seed texture, volume, hardness, seed coat percentage and water absorption of seed, crude protein, crude fat, crude fiber, toal sugars and starch content. Genotype × environment interaction was significant for fresh pod traits such as curvature, length/width and width/thickness ratios, and texture, and for dry seed traits such as length/width and width/thickness ratios, volume, hardness and water absorption. Estimates of heritabilities and constancies (genotypic/environmental variance) were higher than 0.5 and 1.0 respectively for pod curvature, length/width and width/thickness ratios of pod and seed, seed volume and seed water absorption. There were high correlations of pod texture with seed length/width and width-thickness ratios, seed coat percentage, seed water absorption, crude fat, total sugars, starch content and crude fiber. Pod texture could be used as an important parameter for evaluating both seed culinary and nutritional quality. In addition, 17 populations showed promise as parental material in a breeding program for pod and seed culinary quality and protein content. The culinary and nutritional quality potential of the accessions such as PHA-0171, PHA-0253 and PHA-0257 which are known as Alubias de riñón could be the base material in a breeding program to obtain accessions with a good seed quality.     

A geographical approach to identify phosphorus-efficient genotypes among landraces and wild ancestors of common bean

Cultivars of common bean (Phaseolus vulgaris L.) capable of yielding well at low levels of native or added phosphorus (P) are highly desirable in many tropical production systems. The objective of the present study was to identify geographical regions which might be sources of such genotypes. A total of 364 landraces, cultivars and wild genotypes, drawn from a broad geographic range, were divided on the basis of growth habit into four field trials, each comprising two levels of P, stressed and unstressed, on an infertile Andosol in Popayan, Colombia. The regression relationship between grain yield per plant in the presence and in the absence of stress was determined, and each genotype's deviation from this relationship was used as a measure of P-efficiency. There was highly significant variation in efficiency among genotypes in all growth habits, and in climbing beans there were consistent regional differences, superior genotypes being identified with greater frequency among those from Bolivia, West Mexico and South Mexico-West Guatemala. The latter region was promising for prostrate bush genotypes also. Wild beans in general performed relatively poorly; it appears that P-efficiency traits in P. vulgaris have been acquired during or after domestication. These results confirm that genetic differences in P-efficiency exist among common bean genotypes and suggest that these are related to geographic origin. Furthermore, the use of a representative sample of germplasm can help to identify segments of the gene bank that are especially promising as sources of desirable traits. This revised version was published online in July 2006 with corrections to the Cover Date.     

Comparison of recurrent and congruity backcrossing for interracial hybridization in common bean

Summary Two common bean (Phaseolus vulgaris L.) cultivars were used for recurrent backcrossing (i.e., repeated backcrossing to one of the parents) to both parents independently and for congruity backcrossing (i.e., backcrossing alternately to both parents). One cultivar, ICA Pijao of race Mesoamerica, was late maturing, with erect type II growth habit and small seeds. The other was Pinto UI 114 of race Durango, an early maturing cultivar with prostrate type III growth habit and medium seeds. One to three recurrent backcrosses (RBC) to both parents, and two rounds of congruity backcrosses (CBC), were made. Thirty-two random lines from each method along with two parents were compared in a reps-in-set design at two locations (Popayán and Quilichao) in Colombia in 1992–1993.Differences between lines derived through different methods were found for seed yield, 100-seed weight, days to maturity, growth habit, and seed color pattern. The RBC to ICA Pijao produced lines with relatively higher seed yield, smaller seed size, and delayed maturity. Similarly, the RBC to Pinto UI 114 produced lines with reduced yield, medium seed size, and earlier maturity. All lines from one, two, and three RBC to Pinto UI 114 had type III growth habit. While most lines from the RBC to ICA Pijao had erect type II Growth habit, a few type III lines were also found in all the RBC. The lines derived from the CBC mostly possessed yield, maturity, and seed characteristics between those of the two parents and lines derived from RBC to both parents independently. They also had more variation for growth habit. To maximize recombination between and retention of desirable traits from distantly related parents used in interracial crosses, use of the CBC is suggested.     

A rapid screening technique to combine resistance to halo blight and bean common mosaic virus in Phaseolus vulgaris L.

Summary Seven bean lines (Phaseolus vulgaris L.) with differential resistance or susceptibility to race 2 of halo blight (Pseudomonas syringae pv. phaseolicola) and a necrosis-inducing isolate of bean common mosaic virus were inoculated with one or both pathogens in combination, to determine the feasibility of dual screening to identify resistance to both pathogens simultaneously. Dual screening yielded the same results as separate screenings. Neither pathogen affected the disease expression of the other. Simultaneously screening for resistance to both pathogens will shorten the recurrent screening-selection cycle of hybridization programs, and accelerate development of resistant cultivars.Abbreviations BCMV Bean Common Mosaic Virus - cvs Cultivars - HB Halo blight - Inoc. Pt. Inoculation point - NLL Necrotic local lesion - React Reaction - SVN Spreading veinal necrosis, System chloro-Systemic chlorosis - VN Vascular necrosis     

Combining ability for nodulation in common bean (Phaseolus vulgaris L.) genotypes from Andean and Middle American gene pools

Ten F₁'s obtained from crosses among five common bean genotypes of Andean (WAF 15, Mineiro Precoce and Batatinha) and Middle American (BAT 304 and Ouro) gene pools were assessed for their combining abilities for root nodulation with Rhizobium tropici strain CIAT 899. The plants were grown under controlled conditions and evaluated for number of nodules per plant (NN), nodule dry weight (NDW), mean nodule weight (MNW) and plant fresh weight (PFW). The subdivision of the treatment effects on the general (GCA) and specific combining effects (SCA) were performed according to Griffing's diallel analysis method 2, model 1. The analyses of variance and estimates of quadratic components showed that non-additive gene effects were more important in the expression of NN and PFW, whereas additive gene effect was predominant for NDW and MNW. A close association was observed between high number of nodules and GCA. Generally, crosses involving parents of different gene pools yielded hybrids with high positive estimates of SCA for all assessed traits. The genotypes of Andean origin WAF 15 and Mineiro Precoce are the most promising parents for breeding programs to increase NN and NDW in common beans. This revised version was published online in July 2006 with corrections to the Cover Date.     

Growth habit and gene pool effects on inheritance of yield in common bean

Summary To satisfy farmer and consumer preferences, breeding efforts to increase yield potential in common bean must take into account the interrelated effects of growth habit, seed size, maturity, and gene pool on yield expression in segregating populations. To examine the relationships among these traits, a genetic study was conducted to determine the effect of growth habit on yield and seed size in crosses among five bean lines from diverse gene pools. Two parental bean lines had determinate, type I growth habits and large seed size typical of the Neuva Granada-Andean gene pool. Two other lines were tropical Mesoamerican types with type II growth habits and small seed size; and the fifth line, G13625, a landrace of the Jalisco gene pool from the Mexican highlands, had a type IV climbing growth habit and medium seed size. Individual F₂ plants from each cross and parental lines were evaluated for growth habit and yield component traits under high input field conditions. The following season, the evaluations were repeated on random F₃ plants. Of the five parental lines, only G13625 showed significant GCA effects for yield in both the F₂ and F₃ generations. Improved yielding ability of G13625 progeny was associated with an increased expression of climbing bean growth habit traits: guide length, climbing ability, node number on main stem, and plant height. Crosses between Andean x Mesoamerican and Andean x Jalisco genotypes, as well between growth habit type I (Andean x Andean) and between type II (Mesoamerican x Mesoamerican) had very low parent-offspring heritability values for yield. Yield heritability was only significant for crosses between Mesomerican x Jalisco gene pools. An apparent simple genetic control of growth habit modification towards semi-climbing and climbing types is proposed as the major reason for increased yields in these crosses. No genetic linkage between genes controlling growth habit and seed size was detected which might restrict the development of high yielding large-seeded type II lines.     

A factor analysis of plant variables associated with architecture and seed size in dry bean

Summary A knowledge of the biological and functional relationships among traits would be of benefit to plant breeders in the choice of traits to use for increasing efficiency of selection in a breeding program. Principal factor analysis was used to identify biological and functional relationships among dry bean (Phaseolus vulgaris L.) traits in segregating populations of a phenotypic recurrent selection program designed to introduce upright architecture into medium-sized dry bean.Two biological interpretations were assigned to factors extracted in the analysis of four seed/pod traits. The first was a size factor, loaded with seed weight and pod width, and the second a number factor, which was loaded with number of seeds per pod and pod length. Similarly, three factors were extracted from the analysis of fifteen architectural traits. The first factor was interpreted as an elongation factor, and was loaded with internode length measurements, while the second factor was interpreted as a structural factor, and was further divided into two subclasses namely, sturdiness factor and profile factor. The sturdiness factor comprised hypocotyl diameter and hypocotyl length, while the profile factor comprised branch angle and number of basal branches. The third factor which was called a reproductive factor was loaded with pod distribution traits and the number and location of nodes on which pods were borne.     

Rooting of bean leaves and use in germplasm evaluation for common bacterial blight resistance

Summary A simple protocol for leaf rooting in beans (Phaseolus vulgaris L.) was developed and used to investigate the reaction of Xanthomonas campestris pv. phaseoli (Xcp), causal pathogen of common bacterial blight disease, in detached versus attached bean leaves. Trifoliate leaves of different sizes (one-third, two-thirds, and fully expanded), either with or without the pulvinus attached to the petioles, were excised from 20 day-old plants of six bean cultivars/lines. Leaf cuttings were cultured in potting medium and then incubated for 5 to 10 days under transparent polyethylene plastic cover in the greenhouse. Roots were readily initiated along the petioles of the leaf cuttings, whether the pulvinus was present or absent. All leaves which were two-thirds expanded and fully expanded developed roots 5 to 7 days after culture. Eighty to 90 percent of the leaves which were one-third expanded formed roots 8 to 10 days after incubation. Laminae of the rooted leaf cuttings were viable and green during the 2 to 3 months period in culture after removing the plastic cover. The common bacterial blight reactions were similar for inoculated attached leaves, detached rooted leaves (inoculated either after or prior to rooting), and moistened detached leaves incubated without rooting. The latter were only usable for evaluation of the Xcp reaction in growth chamber experiments but not under greenhouse conditions. The rooted leaves would be useful for screening bean lines for multiple disease resistance, especially if the pathogens require different environments for disease expression.Abbreviations CBB Common Bacterial Blight - Xcp Xanthomonas campestris pv. phaseoli (Smith) Dye     

Current status of tropical fruit breeding and genetics for three tropical fruit species cultivated in Japan: pineapple,mango, and papaya

Tropical fruit crops are predominantly produced in tropical and subtropical developing countries, but some are now grown in southern Japan. Pineapple (Ananas comosus), mango (Mangifera indica) and papaya (Carica papaya) are major tropical fruits cultivated in Japan. Modern, well-organized breeding systems have not yet been developed for most tropical fruit species. Most parts of Japan are in the temperate climate zone, but some southern areas such as the Ryukyu Islands, which stretch from Kyushu to Taiwan, are at the northern limits for tropical fruit production without artificial heating. In this review, we describe the current status of tropical fruit breeding, genetics, genomics, and biotechnology of three main tropical fruits (pineapple, mango, and papaya) that are cultivated and consumed in Japan. More than ten new elite cultivars of pineapple have been released with improved fruit quality and suitability for consumption as fresh fruit. New challenges and perspectives for obtaining high fruit quality are discussed in the context of breeding programs for pineapple.     

Inter-and intraracial hybridization and selection for seed yield in early generations of common bean,Phaseolus vulgaris L.

Summary Five parents from each of four race groups of common bean (Phaseolus vulgaris L.) were hybridized to produce five crosses within each group. Also, five crosses were made for each of the six possible combinations among four groups. Parents, F₁ and F₂, and parents, F₂ and F₃ were evaluated for seed yield in 1990 and 1991, respecitively, at two locations in Colombia.Yield of parents belonging to Middle American races and crosses among them was higher than that of races of Andean origin. Positive correlations were found among the mid-parent value, F₁, F₂ and F₃. Also, the mid-parent value predicted the mean seed yield of all possible lines that could be derived from the F-generation in 42 out of 47 crosses. Four crosses, all between common bean races of Andean and Middle American origin, indicated a possible loss of favorable epistatic parental alleles. On average, mean yield of interracial F₁ hybrids was higher than that of intraracial ones.Positive heteroris (26.4%–123.8%) over the mid-parent in 31 crosses, and F₁s yielding higher (23.7%–91.8%) than the high parent in 20 crosses and yielding higher (22.1%–53.2%) than the highest control among all parents (MAM 13) in 12 crosses, were found. Heritability, estimated by the parent-offspring regression, ranged from 0.42± 0.07 to 0.49±0.04. Expected and realized gains from selection (at 20% selection pressure) ranged from 10.3% to 21.0% over the mean of F₁ hybrids and F₂ and F₃ population bulks.     

The introduction of the common bean (Phaseolus vulgaris L.) into Western Europe and the phenotypic variation of dry beans collected in The Netherlands in 1946

The first introduction of common bean from Central/South America into Western Europe most likely took place around 1500. The attractive bean seeds and their easy transportation warranted numerous additional introductions, not only from the Americas, but also from other areas where the common bean had been introduced. Bean seeds also must have been transported all over Europe both locally and internationally. Natural and human selection took place in and among introductions as well as in hybrid populations. Strong human selection may have led to pure garden forms which often occur in The Netherlands. In the period 1945–1948 a dry-bean collection of some 1500 accessions was made in The Netherlands. These have been classified into various criteria: 1. landraces, ‘primitive’ garden forms and cultivars, 2. (semi)climbing or bush types, 3. various seed colour types, including whitish, brownish, yellowish, black and variegated. This revised version was published online in July 2006 with corrections to the Cover Date.